Summary The Heli-Transport Services Inc. Eurocopter AS350 B1 Astar helicopter (registration C-GZCN, serial number 2207) departed a mining camp 176 nm northeast of Chibougamau, at 0800 eastern daylight time en route to a drill site 20 nm to the southeast. Approximately four minutes after departure, the helicopter broke up in flight and descended rapidly to the ground. The pilot, the sole occupant, was fatally injured and the aircraft was destroyed. Ce rapport est galement disponible en franais. Other Factual Information The helicopter C-GZCN was under contract to Melkior Resources Inc. (Melkior), which was engaged in the acquisition and exploration of Canadian mining properties, primarily in Ontario and Quebec. Melkior had positioned an exploration team in the Otish Mountains northeast of Chibougamau, Quebec, to explore for uranium. Heli-Transport Services (Canada) Inc. (Heli-Transport), based in Carp, Ontario, is a helicopter transport company operating under Sections702 and 703of the Canadian Aviation Regulations (CARs), and was contracted to support Melkior's drill site needs, such as crew shift changes, drill site supplies and drill moves. Heli-Transport operates a base in Trois-Rivires, Quebec, where the accident helicopter had undergone extensive maintenance to ready it for the summer season. The helicopter was manufactured in 1989 and imported into Canada in2004. It was operated and maintained by Heli-Transport in accordance with existing regulations and approved maintenance organization (AMO) procedures. All modifications, mandatory airworthiness directives, and required maintenance had been completed. The helicopter's weight and centre of gravity were within the prescribed limits during the flight. The helicopter was not equipped with a flight data recorder or a cockpit voice recorder, nor was either required by regulation. The helicopter had flown approximately 9380 total hours, including 35hours since the completion of the last scheduled major aircraft inspection, which took place between 15March2007 and 14May2007. The helicopter also underwent the 3500-hour inspection on the main gearbox (MGB) epicyclic reduction gear module, main rotor shaft corrosion inspection, and the modification of the rear fuselage structure attachment bulkhead. This work was completed, and the paperwork was signed on 14May2007. During the 3500-hour inspection, the MGB epicyclic reduction gear module was removed from the MGB. This work involved removing the main rotor blades, disconnecting the flight controls, and separating the upper and lower rotor shaft casings of the MGBto gain access to the epicyclic reduction gear module. The epicyclic reduction gear and the rotor shaft casings were sent for overhaul to the Eurocopter Canada facility. They were returned to Heli-Transport on 20April2007, and the epicyclic reduction gear module was re-installed on theMGB. Maintenance activities were completed on 15May2007. The aircraft was ground and flight tested to verify the proper installation of the various system components, to check for fluid leaks and to assess overall aircraft performance. The helicopter was flown for 1.5hours. The main rotor system was checked with a tracking system to ensure the proper flight path and balancing of the rotor blades. During these ground and flight tests, it was noted that the main rotor rpm (Nr)indicated 10revolutions per minute (rpm) lower than the normal 394rpm setting. This lower rpm was still within limits. It was later confirmed through the tracking system that the actual main rotor rpm was not below normal, and that the anomaly was an indicator problem. During one of the test flights, ground resonance1 occurred after landing, while bringing the throttle from full power to idle. Power was immediately re-applied and the resonance disappeared. The aircraft was repositioned without further ground resonance. The maintenance personnel checked all dynamic components as per the maintenance manual and replaced the steel strip vibration absorbers located at the aft end of both skids. In addition, a humming noise was heard during some of the engine ground runs. An engine technical representative present during these engine ground runs indicated that a similar hum had occurred on other AS350Bs. The hum subsequently stopped, and troubleshooting to find the source of the hum also stopped. During another session of ground and flight testing, the main rotor gear box (MGB) chip warning light illuminated. The magnetic chip detector was inspected, and fine metal fuzz was observed. The Eurocopter technical representative was informed. He referred the aircraft maintenance engineers (AMEs) to Maintenance Manual (MM) 05.53.00.6082 for special inspection procedures required after such an event. According to the Eurocopter technical representative and the special inspection procedure documents, the presence of metal fuzz or fine metal particles after replacing parts such as the MGB epicyclic reduction gear module is not abnormal. The special inspection procedure was performed, and no other MGBchip warning light appeared. Each of the above problems was addressed as per the manufacturer's maintenance manual. There were no further indications of problems, and the aircraft was released for flight. The helicopter had been expected at the mining camp in the Otish Mountains on 15 May 2007. Due to the delay in maintenance, the pilot, along with an AME, departed with the helicopter late in the day on 16 May 2007 and arrived on site before noon on 17 May 2007. The pilot was certified and qualified for the flight in accordance with existing regulations. She had approximately 1600hours of total flying time, with 110hours on the EurocopterAS350. She had completed her training and pilot proficiency check in April2007. The AME assigned to the Melkior contract was licensed in2005, and had completed the technical Eurocopter AS350 helicopter course. He had worked for Heli-Transport as an independent contractor on many occasions during the previous three years. Over the 10 days before the accident flight, the pilot inspected the aircraft before the first flight of each day, and the AME inspected it again at the end of the day. The AME continued to monitor the low Nr indication, and had attempted various corrective maintenance measures including many consultations by telephone with the maintenance base in Trois-Rivires in order to rectify the problem. None of the proposed solutions resolved the problem. On 22 May 2007, the MGB chip warning light illuminated. The AME inspected the magnetic chip detector and observed fine metal fuzz. The oil and filter were replaced, and the maintenance manual special inspection procedures were performed before releasing the helicopter for flight. This event was reported to the maintenance base in Trois-Rivires. On 25 May 2007, the pilot experienced ground resonance while landing on a log pad. The resonance immediately dissipated when the pilot reapplied engine power and repositioned the helicopter on the pad. The AME inspected the vibration absorbers and the relevant dynamic components for their condition and their attachments. No abnormalities were found. Other than the above seemingly unrelated problems, the pilot did not report any aircraft performance or other helicopter related problems during this period. On 27 May 2007, the pilot inspected the helicopter for flight and departed at 0645 eastern daylight time3 with a drill crew for the morning crew change. Weather conditions were appropriate for visual flight rules (VFR) flight. At the drill site, the pilot completed three sling moves, and then returned to camp with the exiting night-shift crew. During the return flight, the low rotor rpm warning horn sounded and the pilot reported hearing a low-frequency hum at idle power. A ground check was performed at idle power to attempt to identify the source of the noise. The AME could also hear the hum, but it would disappear as soon as power was increased to 100percent. The aircraft then departed on the accident flight. It was observed after take-off to be en route to the drill site. A trail of what appeared to be smoke was coming from the helicopter, and the helicopter descended rapidly toward the ground. The camp manager and the AME were advised, and a search and rescue plan was initiated. The helicopter was found at approximately 1000in a swamp, 8 nm from the mining camp. The aircraft was partially inverted, and on its right side. The main rotor blades struck the cockpit while in flight, fatally injuring the pilot and severing the cabin roof, sidewalls, and doors. The wreckage debris was spread over a distance of approximately 700feet in a northeast direction. The tail boom was separated from the helicopter and was found approximately 120feet east of the main cabin. The wreckage was transported to the TSB Engineering Laboratory for further examination. During the accident flight, the tailboom had separated from the fuselage at the rear fuselage bulkhead rivet line. The tailboom had recently been modified to fulfill the requirements for Airworthiness Directive (AD) F-2004-035, Fuselage - Rear Structure Junction Frame. The modification of the frame allowed compliance with Alert Service Bulletin05.00.43. This modification consisted of removing the rivets and installing a reinforcement doubler at the rear fuselage bulkhead, along with new rivets. The rear fuselage structure modification and the repair procedures were examined, and a metallurgical examination was completed. No deficiencies were found in the rivets, or in the quality of the riveting. During the in-flight break-up of the helicopter, the tailboom was subjected to a load in excess of its design limits, and the rivets failed in overstress. The helicopter engine (Turbomeca model Arriel 1D, serial number7023) was sent to Turbomeca Canada for examination. Under the supervision of a TSB investigator, the engine was disassembled. All damages found were consistent with an engine that was producing power at impact. The main rotor components were examined at the TSB Engineering Laboratory. The main rotor system had impact marks along the Starflex arms and main rotor blades' attachment bushings. These impact marks indicate that the main rotor was rotating when it struck the forward fuselage. During the post-accident examination, the MGB was split at the main rotor shaft lower casing assembly line. The six main rotor shaft retaining bolts and the self-locking centre bolt attaching the epicyclic reduction gear to the mast were found unscrewed in the bottom of the sun gear. All of the associated locking tabs and bolts were accounted for. Additionally, snap ring segments were found above the phonic wheel (spacer assembly) in the rotor shaft upper casing (see Figure1). Figure1.AS350 Main gearbox schematic The 3500-hour maintenance tasks, including the reinstallation of the epicyclic reduction gear module, were completed using an electronic version of the Eurocopter Maintenance Manual MM63.10.16.4034 as the primary reference (see AppendixA). To perform the reinstallation of the epicyclic reduction gear module, the electronic version of the Maintenance Repair Manual (MRM) work card (WC)62.30.16.7015 (see AppendixB) was referenced via a hyperlink within a caution note (see below) in the MM63.10.16.403. The task headings within the 15pages, including 9diagrams, of the work card were not all applicable to this specific epicyclic reduction gear module reinstallation because there was no need to replace the mast bearing, the main rotor shaft or the rotor mast seal. However, the reinstallation of the main rotor shaft was necessary and reinstallation steps were outlined in WC62.30.16.701. Although several of the reinstallation steps were not necessary, some of the steps were crucial to ensure the proper reinstallation sequence. The AMEs who completed the reinstallation did not follow WC62.30.16.701. Section 3.2 (b) of MM 63.10.16.403, the primary reference, assures the snap ring is fitted, but does not specify the installation sequence. The maintenance manual states the following; "Coat the splines6 with grease and install planet gear carrier on rotor shaft, after checking for presence of snap ring." This instruction is accompanied by the following caution note: These instructions do not specify the installation sequence of the snap ring. They simply state to ensure its presence. However, MRM WC 62.30.16.701, section4, subsection 4.1e) (see AppendixB), does specify the snap ring installation sequence. It states, in part, to install the casings on the rotor shaft and to fit the snap ring, and includes the following caution note: During the final steps of the epicyclic reduction gear module installation, the AMEs used a borescope to visually inspect the interior of the main rotor shaft. To confirm proper installation of the mast retaining bolts, the AMEs confirmed an equal number of visible threads on the mast retaining bolts as per MM63.10.16.403, section3, subsection3.2d) and MRM WC62.30.16.701, section4, subsection4.2g),h), andi). The AMEs saw approximately 1threads showing on all the mast retaining bolts. The instructions do not specify a minimum number of visible threads required to ensure correct installation. During the post-accident examination, the six mast retaining bolts and the self-locking centre bolt attaching the epicyclic reduction gear to the mast were found unscrewed in the bottom of the sun gear. In addition, segments of the snap ring were found above the spacer assembly as opposed to below the spacer assembly. The snap ring's sole function is to retain the spacer assembly, the casings and housings on the main rotor shaft while manipulating the MGB upper module before the epicyclic gear module is bolted on. It is not intended to take either the aircraft weight or dynamic loads experienced in flight. If the snap ring is installed in the wrong sequence, which was the case in this occurrence, it will move out of its groove due to overload forces. This will release the tension on the mast retaining bolts and locking tabs. The tabs will no longer function as locking devices. The direction in which the sun gear turns and rubs the bolt heads allows for the progressive loosening of the mast retaining bolts. When the bolts no longer secure the mast, the main rotor shaft will move vertically and cause the main rotor blades to fly down into the forward fuselage. The AME in charge of the MGB epicyclic disassembly and reassembly project had been employed with Heli-Transport since2002, and had been fully licensed since2004. He completed his technical EurocopterAS350 Astar type training in2005, and had performed a reassembly of the MGBepicyclic reduction gear module only once before. Although he was present for the disassembly and reassembly of the MGB, he was not present during the ground-runs and flight tests. The other licensed AME had been employed with Heli-Transport since June 2004, and had been fully licensed since January 2007. He completed his AS350 type training in October 2006. He had never performed a reassembly of the MGB epicyclic reduction gear module. At different times during the maintenance work, these AMEswere assisted by an AMEapprentice. All maintenance operations are overseen by the production manager. Human factors training was completed by all company AMEs in2005. The work environment, equipment, workload, work and rest periods were examined during the investigation, and were not considered to have negatively affected the work carried out. The AMEs were trained and comfortable working with the English version of Eurocopter's reference documentation. Of the 2832 Eurocopter AS350, AS355, and AC130 helicopters operating with the 4-contact bearing and MGB assembly, as was the case for the occurrence helicopter, 993are operated in Canada and the United States. A similar occurrence happened in Spain in February 1993. The helicopter had been in maintenance following a mast oil leak. The MGB was opened, the seal was changed, and the unit was re-assembled. The helicopter had flown approximately 17hours after maintenance when a grinding noise was reported. The maintenance inspection did not find the cause of the noise. On a subsequent flight, the helicopter went into a dive before crashing and burning. Although the symptoms were similar to this occurrence, they were not linked to the snap ring installation sequence. Following the accident in Spain, Eurocopter issued Telex Service Letter01-41, which addressed the troubleshooting of abnormal noise if maintenance of the mast epicyclic reduction gear assembly had been recently (100hours) performed. Eurocopter also added the mast retaining bolts borescope inspection to the re-installation instructions to ensure the proper installation. Caution notes were added to MM63.10.16.403, section3, subsection3.2b) and MRM WC62.30.16.701, section4, subsection4.1e). Transport Canada (TC) issued Airworthiness Directive (AD) 93-030-065(B) to ensure that Canadian operators comply with the Telex Service Letter issued by Eurocopter. In Canada, a similar occurrence happened in June2006. The snap ring had been installed in the wrong sequence with respect to the phonic wheel/spacer assembly. The helicopter had undergone maintenance at approximately 92.5flying hours before the occurrence. Maintenance personnel had previously completed the 12-year inspection of the main rotor shaft, which required that similar maintenance operations be performed on the MGB. The information gathered indicated that no ground resonance, Nrproblems and/or hum were reported preceding the occurrence. While in flight, the pilot reported a sudden bang, and the MGB warning chip light came on. The pilot landed and had the helicopter inspected. The MGBwas split and it was observed that six of the seven mast retaining bolts had loosened but were still holding the mast. The other bolt was found in the bottom of the sun gear. As with the above occurrences, Nr problems and/or hum and ground resonance were not present even though the snap ring installation sequence was incorrect. Although information from the occurrence in Canada was unofficially disseminated to a few members of the helicopter community, no formal process was in place to identify and inform the community at large. The manufacturer was not made aware of the occurrence or the circumstances leading to it. No specific action or procedures were put in place to prevent a recurrence. TC's present definition7 of a Reportable Service Difficulty found in CAR591 includes human factors-related issues only if the Instructions for Continued Airworthiness (ICA) for the product are determined to be the cause of the error. In Advisory Circular (AC) 591-001, TC maintains that "An organization's Safety Management System should address these occurrences." TC provides a Service Difficulty Reporting Logic Chart (TP14134B) to assist in the determination of defects, failures and malfunctions that require reporting to the Service Difficulty Report (SDR) program (see AppendixC). Heli-Transport's AMEs were aware of the June2006 occurrence and double checked, as per the MM, that the snap ring was in place and, by using the borescope, that all bolts had the same number of threads visible. However, this operation did not guarantee a correct installation of the epicyclic reduction gear assembly.